INDEX

Explanations

phrases related to subscription or engagement with content

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs24576-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.8.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

bane

-0.69

semb

-0.69

wards

-0.69

[/

-0.65

 coord

-0.61

xual

-0.61

ferred

-0.61

wagon

-0.60

 scapego

-0.59

naire

-0.59

POSITIVE LOGITS

 Delicious

0.59

leans

0.59

 POLITICO

0.58

 subscribing

0.57

 Kavanaugh

0.57

 subscribe

0.57

ood

0.56

 reprint

0.56

apps

0.55

 Birch

0.55

Activations Density 0.029%

phrases related to subscription or engagement with content

No Comments

No Known Activations